A Multi-Scale Global Fusion-Based Method for Surface Fissure Extraction from UAV Imagery

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Review of the Manuscript

Title: A Multi-Scale Global Fusion-Based Method for Mining-Area Surface Fissure Extraction

General Assessment

The manuscript proposes a novel semantic segmentation framework (MGF-UNet) for surface fissure extraction in mining areas using UAV imagery. The study integrates multi-scale convolutional feature extraction, attention mechanisms, selective Transformer-based context modeling, FiLM modulation, and frequency-domain feature fusion. The topic is relevant to the fields of remote sensing, geomatics, and geohazard monitoring.

The paper is generally well structured and presents extensive experimental validation on both self-constructed and public datasets. The proposed method shows competitive performance compared with state-of-the-art approaches.

However, several methodological, experimental, and presentation-related issues should be addressed before the manuscript can be considered for publication.

Overall, the work is promising, but major revisions are recommended.

 

1. Major Comment – Title and Scope Revision

The current title emphasizes “Mining-Area Surface Fissure Extraction”. However, the manuscript does not sufficiently analyze or exploit mining-specific geological, structural, or operational factors that control fissure development.

The proposed method is presented and validated mainly as a general fissure/crack segmentation framework based on UAV imagery, and most methodological and experimental discussions are not explicitly linked to mining-induced deformation mechanisms.

Therefore, the scope of the paper appears broader and more generic than suggested by the current title.

It is recommended that the authors revise the title to adopt a more general formulation focused on surface fissure extraction, rather than explicitly on mining areas.

For example, possible alternative titles could be:

• A Multi-Scale Global Fusion-Based Method for Surface Fissure Extraction from UAV Imagery
• A Multi-Scale Feature Fusion Network for Surface Fissure Segmentation in High-Resolution Remote Sensing Images

 

 

2. Moderate Comment – Use of the Term “Ecological Environments” – Line 3

In the Abstract and other sections, the term “ecological environments” is used in a rather generic manner and is not explicitly analysed in the manuscript.

Since no specific ecological indicators, biodiversity assessments, or ecosystem evaluations are presented, the use of this term appears overstated.

The authors are encouraged to replace “ecological environments” with more precise and scientifically appropriate expressions, such as:

• “land surface environment”,
• “natural environment”,
• “surface environment”,
• “environmental conditions”,
• “landscape environment”.

Alternatively, if ecological impacts are to be emphasized, dedicated analyses and references should be provided.

3. Major Comment – Terminology and Conceptual Consistency – Line 4 and in all maintext

Throughout the manuscript, including the Abstract and Introduction, the term “mining area” is repeatedly used, although the proposed method and experimental analysis are not specifically tailored to mining-related geological or geomechanical processes.

It is recommended that the authors adopt more general and technically appropriate terminology, such as:

• “deformation-affected areas”,
• “subsidence-prone areas”,
• “surface deformation areas”,
• or simply “surface fissure-affected areas”,

depending on the intended scope.

This revision would improve conceptual consistency between the stated objectives and the actual methodological contribution.

4. Minor Comment – Line 44

In the sentence referring to Zou et al., the citation is missing at the appropriate position and is instead placed at the end of the sentence.

The reference to Zou et al. [5] should be inserted immediately after the authors’ names and removed from the end of the sentence, in accordance with standard citation practice.

The authors are encouraged to revise the sentence accordingly to improve clarity and consistency of referencing.

5. Minor Comment – Line 51

Throughout the manuscript, there is a consistent formatting issue in which no space is inserted between words and the corresponding reference brackets.

A space should be added between the text and the square brackets enclosing the citation (e.g., “areas [7].” instead of “areas[7].”), in accordance with standard typographical conventions.

The authors are encouraged to revise the manuscript accordingly.

6. Moderate-to-Major Comment – Underestimation of UAV Applicability

The sentence describing the role of UAV remote sensing is overly simplified and underestimates the broad applicability of UAV-based surveys in Earth and environmental sciences.

UAV platforms are currently widely used in several disciplines, including:

• applied geology for landslide mapping and monitoring,
• earthquake geology for surface rupture and fault mapping,
• volcanology for eruptive fissure and fracture analysis,
• geoengineering and structural monitoring for building damage assessment.

The authors are encouraged to expand this statement to better reflect the multidisciplinary relevance and operational versatility of UAV remote sensing in geological and environmental applications.

Such a revision would strengthen the scientific context of the study and improve the positioning of the proposed method within current research trends.

Suggested sentences could be inserted at line 51 after “areas [7].”:

Low-altitude unmanned aerial vehicle (UAV) remote sensing provides high-resolution, flexible, and cost-effective observations and has become a key tool across a wide range of geological and environmental applications. UAV-based surveys are extensively used in applied geology for landslide monitoring and mapping [sugg ref1], in earthquake geology for surface rupture and fault characterization [sugg ref2], in volcanology for the analysis of eruptive fissures and fracture systems [sugg ref3], and in geoengineering for structural damage assessment [sugg ref4]. These capabilities significantly enhance the accuracy and timeliness of fissure detection and provide a reliable data foundation for fine-scale geological hazard identification.

References suggested

[sugg ref1] Cheng, Z.; Gong, W.; Jaboyedoff, M.; Chen, J.; Derron, M.-H.; Zhao, F. Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks. Remote Sens. 2025, 17, 1900. https://doi.org/10.3390/rs17111900

[sugg ref2] Cirillo, D.; Tangari, A.C.; Scarciglia, F.; Lavecchia, G.; Brozzetti, F. UAV-PPK Photogrammetry, GIS, and Soil Analysis to Estimate Long-Term Slip Rates on Active Faults in a Seismic Gap of Northern Calabria (Southern Italy). Remote Sens. 2025, 17, 3366. https://doi.org/10.3390/rs17193366

[sugg ref3] Darmawan, H.; Walter, T.R.; Brotopuspito, K.S.; Nandaka, I.G.M.A. Morphological and structural changes at the Merapi lava dome monitored in 2012–15 using unmanned aerial vehicles (UAVs). Journal of Volcanology and Geothermal Research 2018, 349, 256-267.

[sugg ref4] Zhang, R.; Li, H.; Duan, K.; You, S.; Liu, K.; Wang, F.; Hu, Y. Automatic Detection of Earthquake-Damaged Buildings by Integrating UAV Oblique Photography and Infrared Thermal Imaging. Remote Sens. 2020, 12, 2621. https://doi.org/10.3390/rs12162621

7. Minor Comment – Line 54

Some references are not formatted in accordance with the journal’s author guidelines.

 

In particular, multiple citations should be written in the format “[8, 9]” rather than using alternative or inconsistent styles.

 

The authors are encouraged to revise all in-text citations to ensure full compliance with the journal’s reference formatting requirements.

8. Minor Comment – Line 70

In several instances, authors’ names are mentioned in the text without the corresponding reference number being provided immediately afterwards.

According to standard citation practice, when authors are cited in the narrative, the related reference should be inserted directly after their names (e.g., “Zou et al. [14]”).

9. Minor Comment – Line 113

The name of the dataset is reported inconsistently throughout the manuscript (e.g., “Mining-area Fissure Dataset” and “Mining-area Surface Fissure Dataset”).

The authors should select a single, consistent designation and use it uniformly throughout the entire text, including figures, tables, and captions, to avoid confusion

10. Major Comment – Lines 127 - 133

In several parts of the manuscript, statements related to geological setting, stratigraphy, and mining activities are presented without any supporting bibliographic references.

Given the extensive existing literature on stratigraphy, mining extraction methods, and subsidence-related deformation, appropriate geological and mining studies should be cited to support these claims.

The authors are encouraged to:

• Provide relevant references on the stratigraphic framework and geological context of the study area,
• Cite studies related to mining extraction techniques and their geological impacts,
• Strengthen the scientific basis of these statements through adequate literature support.

This revision is necessary to improve the credibility and academic rigor of the manuscript.

11. Minor Comment – Line 138

For clarity and scientific accuracy, the authors should explicitly indicate the year (or time period) to which these references apply.

12. Minor Comment – Line 141

In reality, it describes the Digital Elevation Model (DEM).

13. Minor Comment – Caption of Figure 1

write: Digital Elevation Model

14. Minor Comment – Line 152

please insert or cite link of the software.

15. Minor Comment – Caption of Figure 2

Write out in full “Mining-area Fissure Dataset” or “Mining-area Surface Fissure Dataset” and then provide the acronym (MFD), and road fissure dataset (RFD)

 

16. Minor Comment – Line 178-179

This section is crucial for the reproducibility of the techniques used, which will also benefit other researchers who wish to replicate the application. I suggest providing a more detailed description of the procedures.

17. Minor Comment – Figure 10

This figure lacks a geographical reference, so I suggest adding a clearly visible North Arrow, perhaps with a white or gray background to make it stand out. Additionally, it would be helpful to add coordinates along the edges of the figure. The red text is hard to read, so consider changing the color or adding a background. Furthermore, the figure caption should specify the source of the image, whether it is derived from the DOMs or from satellite data.

18. Minor Comment – Line 521

please specify the number of the table.

 

19. Major Comment – Expansion and Strengthening of Section 6.2 (Discussion and Future Perspectives)

Section 6.2 mainly focuses on the limitations of the proposed approach, while its practical advantages and its potential role within a broader geological hazard assessment framework are only marginally discussed.

It is recommended that the authors substantially expand this section by:

1. More clearly highlighting the advantages of the proposed method, particularly in terms of:
• spatial resolution,
• operational flexibility,
• cost-effectiveness,
• capability for rapid post-event surveys.
2. Providing a critical comparison with previous studies that have employed UAV-based remote sensing for mapping:
• landslide-related fractures,
• fault surface ruptures,
• volcano-tectonic fissures.

Relevant literature in these domains should be discussed to better contextualize the contribution of the proposed method.

1. Emphasizing the importance of multidisciplinary integration.

In addition to optical and thermal infrared data, the authors are encouraged to discuss the integration of geophysical surveys (e.g., electrical resistivity tomography, GPR, seismic methods) to investigate subsurface fracture systems and deformation patterns.

A combined remote sensing–geophysical framework would enable a more comprehensive understanding of fissure development and significantly improve geological hazard monitoring and early warning capabilities.

Specific Comment on the Proposed Future Research Paragraph (Lines 578–592)

The following paragraph could be further strengthened by explicitly incorporating a multidisciplinary perspective and by highlighting the added value of the proposed approach within integrated hazard assessment frameworks.

The authors are encouraged to revise and expand the paragraph as follows (suggested direction):

In addition to multi-source remote sensing fusion, future research should consider the integration of geophysical observations to investigate subsurface fracture systems and deformation processes. Combining UAV-based optical and thermal imagery with geophysical measurements would provide complementary surface and subsurface information, enabling a more comprehensive characterization of fissure development mechanisms. Such a multidisciplinary monitoring strategy has strong potential to enhance geological hazard assessment, early warning, and risk mitigation in complex environments.

Furthermore, explicit references to existing UAV-based studies on landslide fractures, fault fractures evidence, and volcano-tectonic structures should be incorporated to support this discussion.

 

20. Ensure that the manuscript has not been previously submitted to or published in other journals. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5616304

 

General Comment – Final Remarks

The authors are strongly encouraged to carefully consider and address all the comments provided in this review and in the annotated PDF.

After appropriate revision and improvement, the manuscript is expected to be significantly strengthened and may become suitable for publication in the journal.

Comments for author File: Comments.pdf

Author Response

Please refer to Response 1 for the detailed revisions. Thank you.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper addresses the significant challenge of accurately extracting elongated and scale-varying surface fissures in mining areas by proposing a novel semantic segmentation network, MGF-UNet. The methodology is technically sound, integrating multi-scale feature sensing(MFS), efficient multi-scale attention(EMA), and a Token-Selective Context Transformer (TSCT) to effectively balance local detail preservation with global context modeling. A particular strength lies in the construction of two specialized datasets (MFD and RFD), which provide solid data support for the experimental validation. The results demonstrate that the proposed method outperforms several mainstream models in terms of accuracy, Dice score, and IoU, while maintaining favorable computational efficiency and strong generalization capabilities across different scenarios.

The paper presents a valuable and well-executed study; however, addressing the points regarding environmental robustness would further strengthen the practical impact of the proposed MGF-UNet.

Weaknesses

• Environmental Robustness: The study area and datasets primarily cover summer conditions with specific vegetation heights; the model's performance under more extreme environmental interferences (e.g., snow cover or different lighting conditions) is not fully discussed.
• Complex Background Analysis: Although the AFF module aims to suppress noise, more qualitative analysis is needed to show how the model handles specific false positives like linear shadows that closely mimic fissures.
• In Section 5.5, there is a citation error regarding the ablation experiments; for instance, the text states '...were conducted on the MFD dataset, as summarized in Table ??', where the table number is missing.
• More lastest research should be considered for related work. e.g. “SCAFNet: A Semantic Compensated Adaptive Fusion Network for Remote Sensing Images Change Detection”.

Author Response

Please refer to Response 1 for the detailed revisions. Thank you.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Editor and Dear authors,

I have carefully reviewed the revised version of the manuscript entitled:

“A Multi-Scale Global Fusion-Based Method for Mining-Area Surface Fissure Extraction” (Manuscript ID: sensors-4151203), submitted to Sensors by Zhou et al.

The authors have adequately addressed all the comments and suggestions raised during the review process. The manuscript has been substantially improved in terms of scientific clarity, methodological rigor, presentation quality, and consistency.

In particular, the revisions have strengthened the discussion, improved the terminology, enhanced reproducibility, and clarified several methodological and contextual aspects that were previously unclear.

In light of these improvements, the manuscript now meets the scientific and editorial standards of the journal and can be considered suitable for publication.

I therefore recommend acceptance of the revised manuscript.

Sincerely,
The reviewer

Author Response

Thank you for your comments, suggestions and help on this manuscript. In addition to the revisions to the original manuscript, the new manuscript has also optimized some of the charts, for example, the border of Figure 7 has been removed.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have carefully revised their manuscript according to my comments and suggestions.  However， there are still the following issues：

1、The format of the article is confused, which affects the reading experience. At the same time, there are some grammatical problems.

2、The author does not clearly explain the motivation behind this manuscript. What are the existing problems and why they are crucial should be explained in more detail. I suggest the author further strengthen the relevant parts of the Introduction.

3、More lastest research should be considered for related work. e.g.  “ORSI Salient Object Detection via Progressive Interaction and Saliency-Guided Enhancement”.

Author Response

Please see the documentation for detailed revisions and replies. Thank you.

Author Response File: Author Response.pdf